MULTI-NEURAL NETWORK MODEL FOR HANDLING REAL-TIME APPLICATIONS USING EMBEDDED HADOOP CLUSTER
S. P. Karthikeyan, Hari Seetha · Journal of Critical Reviews · 2020
Data analytics is a key feature and requirement for the growth of any industry or organization. However, todays small-scale industries with low budgets and big data often face relinquished profit issues, which can only be solved by a combination of big data warehouse systems and efficient data analytics tools. The paper considers the problem posted due to lack of data analytics attached to big database structures or efficient data mining and analytics tools with the absence of effective databases. This paper looks forward to developing a working idea on embedded Hadoop with predictive and statistical data analytic tools like the Artificial, Convolutional and Generative neural networks and machine learning algorithms like Gradient boosting (xgboost), support vector machine (SVM) and regularization regression models. The research takes into consideration the harsh and noise persisting data and use multilevel neural networks to solve this problem. This paper made use of blob segmentation and detection for better identification of hotter regions.